Fibril-forming collagens represent an evolutionary related group of structurally similar molecules within the larger family of collagen proteins. Characterization of naturally occurring mutations has provided a model whereby clinically distinct phenotypes are predicted on the basis of how specific mutations alter normal fibrillogenesis. This model, originally derived from studies of type I collagen defects in osteogenesis imperfecta and Ehlers Danlos syndrome type VII, has been modified and extended by recent correlations of type II collagen defects with several chondrodysplasias and of type III collagen defects with Ehlers Danlos syndrome type IV. From analysis of the skeletal dysplasias, the pathogenic role of fibrillar collagen defects in more common clinical entities has been suggested and awaits rigorous proof. Although informative, these collective studies have revealed important exceptions to predictions of the original pathobiochemical paradigm, and, thus, they have initiated a more rigorous reconsideration of the deductive model. As an alternative, investigations are currently underway to generate transgenic mouse models of human collagenopathies. This task will not only clarify the complexity of collagen pathophysiology, but it will also permit the development of therapeutic strategies.